I’ve never been totally sold on the argument that Denuvo anti-tamper DRM has a big effect on performance. I’ve played plenty of games that feature Denuvo and happen to run great, so I discounted it as a fallacy. As Denuvo becomes more and more of a hot topic however, I wondered whether there was anything that could be done to settle this argument once and for all. To answer the all-important question - does Denuvo affect PC performance.
Now, saying this and doing it are two different matters entirely. Either a game has Denuvo or it doesn’t, meaning there’s no real way to compare performance. There is one that’s a little different though. DOOM. Bethesda published DOOM in May 2016. It ran on the id Tech 6 engine with OpenGL and, shortly after launch, Vulkan support. It was also protected by Denuvo. The key word being ‘was’, because id Software patched Denuvo out of DOOM in December 2016. Boot up DOOM now and it’s Denuvo-free.
What’s cool about this is that we extensively benchmarked DOOM at launch, while reviewing graphics cards, and when Vulkan support was added. We have a very clear idea of how DOOM ran before Denuvo was removed. So as an experiment it seems logical to test DOOM now, after Denuvo has been removed, and we can see whether performance has indeed got better without the notorious DRM.
A quick disclaimer before we get into this. A lot has changed since we did our GTX 1060 Vulkan benchmarks in July 2016. There has been literally dozens of new Nvidia drivers, any of which may have a positive (or negative) impact on DOOM’s performance. DOOM itself has also been patched, another factor which may affect performance. To that end, we do not know for sure the results of this benchmark are directly indicative of Denuvo’s performance impact, but they do at least form some sort of guideline.
Aside from that, everything remains exactly the same for these benchmarks. It’s the same benchmark run, on the same level, performed three times at each setting, with the exact same hardware. Under the hood we have an MSI GeForce GTX 1060 Gaming X 6G, an Intel Core i7-5820K CPU and 16GB DDR4 memory.
DOOM with Denuvo vs DOOM Post-Denuvo - GeForce GTX 1060 6GB, Intel Core i7-5820K, 16GB DDR4 RAM
Well, these results leave us none the wiser. My initial 1080p benchmarks seemed fairly conclusive - Denuvo DRM causes a 6-7% decrease in frame rates. However, at 1440p there was nothing between the two. I tested it multiple times to double check and sure enough, there's no impact. However I just can't ignore that 6% frame rate cost at 1080p. It could be down to anything; new drivers; a patch; even that it's cooler in the testing room today; or it could be that Denuvo is indeed having a noticeable performance impact.
There's also question marks hanging over the CPU. The intel 5820K is a bit of a beast, it's comfortable handling high loads. Denuvo allegedly puts a repeated strain on both CPU and RAM, so there is potential the difference could be even more noticeable on a system with a weaker processor. Maybe then Denuvo does actually cause a crippling drop in frame rate. Unfortunately without results from before Denuvo was removed, we can't test it effectively.
Coming from a position where I didn't necessarily believe Denuvo had a noticeable performance impact, this is has certainly opened my eyes to the possibility it could on certain machines. There's some big question marks raised over the veracity of these results, but taking them at a face value there is an impact, and if this impact is proportionately larger with weaker hardware, it's easier to see some more noticeable frame rate drops happening.
Again, we can neither confirm nor deny Denuvo DRM affects PC game performance, but the results are what they are. What do you make of them? And have you noticed any sure-fire times when Denuvo has affect your PC gaming performance? Let us know!